Pressure gradient flow friction prediction method based on long and short term memory network

The invention discloses a pressure gradient flow friction prediction method based on a long-short term memory network, and relates to the field of shock wave-boundary layer interference flow prediction.The method comprises the steps that S1, physical quantity parameters capable of influencing pressure gradient flow friction prediction are selected from data parameters obtained in ground wind tunnel test capacity or numerical calculation; obtaining a flow field local Reynolds number Rex, a Mach number Ma infinity, a near-wall flow deflection angle phi, a wall surface local curvature radius Cw and a local wall surface pressure pw; s2, performing normalization processing on the physical quantity parameters in the S1 to obtain corresponding normalized variables; and S3, inputting the normalized variable obtained in the step S2 into a long-short term memory network to complete local friction prediction. The invention provides a pressure gradient flow friction resistance prediction method based on a long and short